Mr. W. S. Kent on the Embryology of Sponges. 149 



tlie same time he delineates an apical aperture and central 

 cavity, the latter lined with a separate layer of so-called en- 

 dodermal cells^ the existence of neither of which is any longer 

 maintained. Still more direct testimony, if needed, in de- 

 monstration of the identity of the constituent elements of the 

 upper and lower portions of the sponge-embryo, even where 

 those of the latter one are of considerably larger size, is af- 

 forded by Plate VI. fig. 16, in which, as will be seen, the 

 cellular elements of the lower jjortion exhibit all the characters 

 of the adult collar-bearing zooids or units, while those of the 

 upper part have arrived only at the semideveloped uniflagellate 

 and collarless condition. This interesting example was met 

 with in a calcareous sponge-form common on the Jersey coast, 

 closely allied to Haeckel's Ascatidra pinus, and having associ- 

 ated with it innumerable other embryos presenting the typical 

 ovate and uniform character delineated in fig. 12. This some- 

 what abnormal example last described furnishes a complete 

 key to the commonly occurring form delineated at fig. 15, 

 this latter, indeed, representing a slight modification of the 

 same type, in which the zooids of the lower portion have still 

 further outstripped their antipodal companions in the race, 

 losing their collars and flagella, and assuming the passive 

 amoeboid state accompanied by a syncytial exudation before 

 these others have so much as developed the first-named 

 structures. Why, in some instances but not in others, this 

 disparity in the degree of growth should exist between the 

 separate units or zooids of the anterior and posterior portions 

 of the aggregate mass is easily explained. On making a 

 suitable section through a sponge-body containing these em- 

 bryos it will be found that in some cases these bodies are 

 released from their syncytial matrix in their entirety, the 

 zooids under these circumstances developing evenly throughout 

 the periphery, while in others they for a while remain partially 

 immersed within the same. In this latter case the zooids of 

 tlie two opposite portions naturally develop at a different rate, 

 those appertaining to the immersed one being temporarily 

 retarded in their growth. In many instances indeed it would 

 seem that the most posterior or deeply immersed cellular con- 

 stituents do not perfect their final subdivision and development 

 into the typical collar-bearing monads until the permanent 

 attachment of the embryo. Mr. Carter has applied to these 

 occasional larger cells at the posterior extremity of the ciliated 

 embryo the title of root-cells, these same, when present, taken 

 collectively, representing the region by which attachment to 

 the selected fulcrum of support is most usually eftected. It is 

 a significant fact that, in cutting open or otherwise examining 



